General Question Orbiter planning vs. Trajectory Browser

Darkriser

Member
Joined
Oct 17, 2019
Messages
12
Reaction score
1
Points
18
Hi All,

A newbie calls for help....

I've already seen many tutorial videos (including the marvelous series from Blixel, Dimitris, etc...) and was
able to reproduce almost everything using XR-2 and XR-5.
Now, I'd like to start creating my own goals/missions using the knowledge I have.

Ok, let's have one of the most simple cases - Earth to Mars.
But (unlike the video tutorials) this time we'll use some real-world trajectories for planning.
Let's find the most optimal trajectory and try to plan this in Orbiter (I use XR-2 or XR-5 for all my attempts).

And just for fun let's plan the journey in both - TransX and IMFD - and we'll see.
IMFD - this one is easy:
  • go to Target Intercept
  • set departure to MJD 60584 (1-Oct-2024) according to the link above
  • set arrival to MJD 60920 (2-Sep-2025)
...and we're done as I'm not interested in launch, eject, etc. at this moment....
TransX - very easy, too:
  • Stage 1 - Escape
  • Stage 2 - Mars (Stage 3 - Encounter, but useless for now)
  • In Stage 2 - set the Eject Date to 60584 and all the velocities to Auto-Min
....done again, seems so trivial.
The result can be seen in the pictures attached.
Mars_IMFD.pngMars_TransX.png

But now the doubts step in...
From David's/Dimitris'es tutorials I know that the most dV efficient trajectory to Mars requires only around dV 2.8k.
My actual trajectory was selected by a tool from NASA as the most efficient in terms of dV!
Why do I see much higher numbers in TransX and IMFD ??
Even when I try shifting the departure/arrival time/date by several days, I'm unable to find anything more efficient.

Why NASA's trajectory browser doesn't match the calculations in Orbiter?
How do I plan some trajectories (e.g. to other planets) in a more advanced way (without just 'guessing' the launch window and
trying to randomly/blindly shift the dates until I get the 'lowest dV')?

Many thanks in advance...
Marcel
 

jedidia

shoemaker without legs
Addon Developer
Joined
Mar 19, 2008
Messages
10,386
Reaction score
1,631
Points
203
Location
between the planets
From David's/Dimitris'es tutorials I know that the most dV efficient trajectory to Mars requires only around dV 2.8k.
I'm not familiar with the tutorial, and with your setup here. Is it on the same date? Because it's totally possible that there was a point in the past where a more optimal trajectory existed.
Also, I'm assuming you're starting from orbit, but... what plane do you start from? Was the injection burn in the tutorials an off-plane ejection that it include a midways injection burn to align the planes? Because then you'd have to add that DV on top of your 2.8k...

Why NASA's trajectory browser doesn't match the calculations in Orbiter?
But... it matches your transX trajectory almost to a T. IMFD has a lot more automation built in, so it's very well possible that it didn't find the optimal solution, but the TransX trajectory you set up manually matches the duration perfectly and the DV is off by only 0.3 km/seconds. You can't really expect it to match exactly, Orbiter isn't reality(TM) and TransX isn't a NASA department, so I'd consider that result pretty impressive.
 
Last edited:

Darkriser

Member
Joined
Oct 17, 2019
Messages
12
Reaction score
1
Points
18
Also, I'm assuming you're starting from orbit
Not at all, I'm still sitting at KSC RWY 15, current date/time and planning my Earth to Mars trajectory.
The Launch and Eject Burn will be planned a bit later, once I know the trajectory.
This is how it's supposed to be done, isn't it?

But... it matches your transX trajectory almost to a T. IMFD has a lot more automation built in, so it's very well possible that it didn't find the optimal solution, but the TransX trajectory you set up manually matches the duration perfectly and the DV is off by only 0.3 km/seconds. You can't really expect it to match exactly, Orbiter isn't reality(TM) and TransX isn't a NASA department, so I'd consider that result pretty impressinve.
This is true.
But by 'doesn't match' I meant that I was looking for the most efficient (in terms of dV) transfer....and Trajectory Browser did find something that looks very close to Hohmann transfer, which should be very efficient.
In Orbiter, however, we're able (if we try really hard) to achieve a trajectory with only dV 2.8k to Mars.
I expected that if I enter the 'most efficient' trajectory (according to NASA) into Orbiter (doesn't matter which MFD) it will result in something very close to dV 2.8k.
This is what I didn't expect - I ended up with dV being (relatively) much more than that.
 

jedidia

shoemaker without legs
Addon Developer
Joined
Mar 19, 2008
Messages
10,386
Reaction score
1,631
Points
203
Location
between the planets
Not at all, I'm still sitting at KSC RWY 15, current date/time and planning my Earth to Mars trajectory.
So the injection burn is calculated directly from the surface? Then a lot of it would depend on how much atmospheric influence during the launch phase is factored in (and that's a pretty large chunk). I can't really make any statements on how either the Orbiter tools or the Nasa browser factor that in. For Orbiter, you could find out by flying the trajectory with 2.8k and measuring how much DV you actually used after the burn, and for the Nasa browser there's presumably some documentation around of the models they're using.

Even with that, keep in mind that atmospheric modeling in Orbiter isn't too precise (that stuff be complicated), and that it even depends a lot on stuff like efficiency of the engine when not in a vacuum, aerodynamics, etc, so even the flight dynamics coded into the vessel module itself have an influence here.
 

BrianJ

Addon Developer
Addon Developer
Joined
Apr 19, 2008
Messages
1,519
Reaction score
640
Points
128
Location
Code 347
I expected that if I enter the 'most efficient' trajectory (according to NASA) into Orbiter (doesn't matter which MFD) it will result in something very close to dV 2.8k.
This is what I didn't expect - I ended up with dV being (relatively) much more than that.
Hi,
a dV of 2.8k to Mars seems pretty low to me (whichever way you're defining it) so I'd be interested to know what eject and arrival dates that solution uses.

Also, a quick look at the NASA trajectory browser page you linked to, shows that:
1) it's using an additional post injection burn of ~570m/s (no other info about the burn though), which neither IMFD or TransX will be using (unless you use IMFD "Target Plane" arrival constraint mode, which gives a similar post-injection burn dV requirement).
2) it shows the "Injection dV" from Earth LEO (200km alt., no other info), whereas (I think) IMFD shows the excess velocity("Vinf") required, possibly the same for TransX. For IMFD the actual dV from LEO will be greater (see IMFD Orbit Eject function).
3) it must be using some kind of "arrival-condition constraint", since changing the trajectory mode to "Flyby" changes the solution.

I guess I'm just trying to point out that the figures given by NASA Trajectory Browser and IMFD or TransX may not be directly comparable, and there may be "hidden" assumptions or constraints used in the trajectory planning. Also NASA TB has a fairly limited range of dates. Also, it makes quite a difference whether you intercept Mars at perihelion or aphelion, etc.etc.etc.

Anyhow, I hope you have fun planning (and making) your trip to Mars!
Cheers,
Brian
 

dgatsoulis

ele2png user
Joined
Dec 2, 2009
Messages
1,859
Reaction score
175
Points
78
Location
Sparta
As already mentioned in this thread, the oV value in IMFD's course plan page and the Total Delta V in TransX's Eject plan stage, refer to the hyperbolic excess velocity aka V infinity (V∞). This is not the same as the injection delta_V, which is used by the NASA trajectory browser (calculated from a 200x200 km parking orbit). In order to be able to compare the plans, one must convert from V∞ to ΔV_inj and vice versa.
Here are some useful equations:
CodeCogsEqn.png
For Earth, we have an orbital velocity of 7.79 km/s for a 200 km parking orbit. Here is a graph of the relation between the hyperbolic excess velocity and the injection deltaV from that parking orbit, ranging from V∞ = 0 = (escape velocity) up to 5 km/s
Untitled-1.jpg
One takeaway from this, is that a big change in V∞ doesn't cost that much in injection dV. For example, to go from 3 km/s to 4 km/s in V∞, we only need ~0.3 km/s more in injection deltaV.

NASA's 3.68 km/s trajectory refers to a V∞ of ~3.2 km/s which is close to what the OP got in both IMFD and TransX.

But now the doubts step in...
From David's/Dimitris'es tutorials I know that the most dV efficient trajectory to Mars requires only around dV 2.8k.
My actual trajectory was selected by a tool from NASA as the most efficient in terms of dV!
Why do I see much higher numbers in TransX and IMFD ??
Even when I try shifting the departure/arrival time/date by several days, I'm unable to find anything more efficient.
It is the most efficient within the constraints you gave it, mainly the date range. To get something within the range you mention (V∞ ~ 2.8 km/s = 3.57 km/s in NASA's browser) you need to extend your search and switch the "Mission type" to flyby. Here is the optimal trajectory in terms of deltaV, with a 1 year flight time constraint :

Untitled-2.jpg

And here is an example that saves almost 100 days of flight time, just by using 20 m/s more on the injection delta-v (As you can see, efficiency is a relative term).

Untitled-3.jpg

How do I plan some trajectories (e.g. to other planets) in a more advanced way (without just 'guessing' the launch window and
trying to randomly/blindly shift the dates until I get the 'lowest dV')?

NASA's trajectory browser is pretty good but the time constraints are a bit limited (up to 2040).
If you want some alternatives, I recommend Piper's Trajectory Planner (body to body, direct), or for something more advanced, Arrowstar's Trajectory Optimizer Tool V2.1.
IIRC that last one can even find slingshots or even multi revolution trajectories (aka go-arounds).

Do take the time to read the manuals, as both tools have a bit of a learning curve in order to use them efficiently.

All the best,
Dimitris
 
Top